AI in Belgium tech. Latest at 9:09 am CET.
Hoş geldiniz, Amedspor! The Süper Lig club from Diyarbakır joins MyGamePlan. Their staff will prepare every opponent and review every match with MGPT, the AI assistant, with the video behind each finding. More on the partnership 👇
Your next customer might not be human. I made a short film about it 🎞️ 𝑻𝒉𝒆 𝒆𝒏𝒅 𝒐𝒇 𝒎𝒂𝒓𝒌𝒆𝒕𝒆𝒆𝒓𝒔? 🙃 𝑶𝒓 𝒕𝒉𝒆 𝒔𝒕𝒂𝒓𝒕 𝒐𝒇 𝒕𝒉𝒆 𝒄𝒉𝒂𝒑𝒕𝒆𝒓 𝒂𝒉𝒆𝒂𝒅!
We just added Grok and DeepSeek to Rankshift. 11 AI models tracked now. A question I hear in almost every demo: do I really need to monitor all of them? Short answer: no. Monitor the ones your buyers use. The catch is that you cannot know which ones matter for your category until you have measured it, and the differences between models are bigger than most people expect. A brand that is cited consistently in Perplexity can be absent in Gemini, and the sources each model trusts differ too. That is why we keep adding models instead of picking a few. You decide which ones run in each scenario, we make sure the data is there when you need it. Grok (xAI) and DeepSeek are available now, next to ChatGPT, Gemini, Google AI Overviews, Google AI Mode, Claude, Perplexity, Copilot, Mistral and Meta AI.
New data from 6sense’s 2026 Buyer Study strongly validates our stance on B2B "AI Visibility." lnkd.in/eM-kbJY6 How much does the chatbot really decide? linkedin.com/pulse/how-much-does-chatbot-really-decide-koen-de-witte-marle?trackingId=uwPwzNRXKpO8Tt2MH1Z1DQ%3D%3D
VESPER BENCHMARKS, EPISODE 1 First up: LIBERO. We scored 100% with zero human demonstrations. LIBERO is typically used to evaluate methods trained on demonstrations, the top models learn from 50 per task. We only gave Vesper the task: the scene, the instruction and the check for the goal. The Vesper agent wrote each program on the first starting position. It then ran unchanged on 10 others, with the objects in different places, and passed every time. No human wrote any code, it all ran in simulation. With the human out of the loop, this scales massively. Solving LIBERO using demonstrations isn't wow in 2026, but doing it without any is a feat. LIBERO-PRO is next. Four clips and the full story → lnkd.in/eMV6abEG
Live from 6sense breakthrough....My latest Blog Post on LLMs. #AI #SEO #Visibility How much does the chatbot really decide? linkedin.com/pulse/how-much-does-chatbot-really-decide-koen-de-witte-marle?trackingId=F%2BuzssSdgXBE%2F12PgG4qHg%3D%3D
++ AI-event AIIA brengt leiders samen rond maatschappelijke impact AI ++ We hebben het vaak over de risico's van AI: jobverlies, privacy, deepfakes en de impact op ons energieverbruik. Maar welke maatschappelijke meerwaarde kan AI creëren? Met AIIA - AI for impact - willen CityCubes en vijf founding partners het debat verbreden en concrete Belgische AI-toepassingen in de kijker zetten. Op 15 april 2027 willen ze in de Corda Arena in Hasselt 1.200 bedrijfsleiders, beleidsmakers en AI-experten samenbrengen voor een dag vol inspiratie, uitwisseling en praktijkvoorbeelden. Met onder meer Peter Hinssen, Steven Van Belleghem en Zuhal Demir op de sprekerslijst. Op het programma staan rondetafelgesprekken, keynotes, een hackathon en een live AI-Championship. Tijdens de avondshow worden ook de allereerste 12 AIIA Awards uitgereikt aan Belgische AI-toepassingen met een positieve maatschappelijke impact. De Sustainable Development Goals vormen daarbij een belangrijk beoordelingskader. Veel succes Dieter Veulemans! Bij #ZigZagHR zijn we fan, maar dat wist je natuurlijk al... >> LEES MEER: lnkd.in/eXXUZC_B
Mac Studio is een enorm krachtige pro-desktop voor alles van AI-inferentie op het device tot het verwerken van gigantische datasets. Mac Studio met M5 Max versnelt complexe workflows, en met de vier Thunderbolt 5-poorten met snelle aansluitmogelijkheden en uitbreidingsopties kunnen teams een optimale configuratie uitwerken. Meer informatie lnkd.in/erqyhd7A #Assign
If AI could save you five hours of work every week, what would you spend that time on? A few weeks ago, I attended a session on the social impact of AI, organised by The Shift vzw/asbl and Vlerick Business School in Brussels. It made me look at AI transformation projects a little differently. We usually approach these projects from a practical angle: Where can we save time? What can we automate? Which processes can we improve? All important questions. But efficiency alone doesn’t tell us whether a transformation is successful. What if those hours could be spent sitting down with the people you’re trying to help? Improving programmes? Building stronger partnerships? Or simply doing more of the work that made you join the organisation in the first place? That’s the thinking behind a Purpose Audit, a short exercise we’ve recently added to our AI transformation projects. Before looking at processes and technology, we ask teams what parts of their work contribute most to their mission, and what they’d like to do more of if they had the time. It’s a small exercise, but it changes the conversation. Instead of focusing only on what AI can take off people’s plates, we start thinking about what it can make possible. Because saving five hours a week isn’t the goal. What you do with those five hours is. And that’s one of the conversations I hope we’ll have at NGO of the Future: how can we make sure AI doesn’t just make NGOs more efficient, but actually helps them deliver more impact?
The world of #AI is now. #ZeroTouchAccounting
Your most valuable data is rarely in a database. It’s in reports, memos & briefings, written for a printer. You can index it, you can embed it; you still can’t add it up. On 11 November, Leon Meijer (GenAI Architect, Ministerie van Defensie) shows how to fix that: a cognitive layer (Indentia BV) on top of Red Hat OpenShift that gives the words in a document an explicit meaning. A ship’s name in a report & a row in a register can then be the same object. The result is a “Corporate Second Brain”: one queryable memory for search, AI agents & applications, air-gapped where it has to be. Our Tech Tribes colleagues built the OpenShift platform this runs on. At Kangaroot, we build & run OpenShift AI platforms too, as a Red Hat Premier Partner: open source, under your control, with 24/7 support. From Made for the Printer via Data Fusion to Made for Agents, 12:15 on the mainstage at Red Hat Summit: Connect Utrecht next week! Drop by our booth afterwards for a conversation. Does your AI run on your own infrastructure today, or on someone else’s?
Business teams experimenteren steeds meer met vibe coding, en vroeg of laat komt de vraag bij IT: Wie beheert dit? Wie beveiligt dit? Wie past er aan als dit nodig is? Waar liggen de verantwoordelijkheden? Op donderdag 10 december brengen we in Gent een kleine groep IT leiders samen rond deze topics. Samen met Mendix overlopen we de volledige levenscyclus van een applicatie. Van eerste idee tot onderhoud. Daarbij laten we zien hoe wij vandaag AI gebruiken om zélf het verschil te maken. Achteraf voorzien we ook tijd en ruimte om ervaringen en ideeën uit te wisselen met collega's die voor dezelfde keuzes staan. Er zijn nog een aantal plaatsen: lnkd.in/eAjyZkma Schrijf u snel in! Mendify Inner Circle: AI in elke fase van applicatie ontwikkeling - Mendify mendify.eu/events/mendify-inner-circle
AI is becoming a new discovery channel for mobile apps! We recently surveyed 1,000 US adults and found that 10% discovered their most recently downloaded app through an AI assistant, almost as many as through Google! But here's what's interesting: AI recommendations don't replace the App Store or Google Play. Users still check ratings, reviews, and screenshots before downloading. AI visibility and ASO are becoming two sides of the same coin. Today, our very own Simon Thillay will join David Barnard on Sub Club Live to discuss what this means for app marketers. 📅 Today, November 5, at 6 p.m. CET. The way people discover apps is changing. Are you ready? lnkd.in/eAA8S-V5 ASO for AI search: how apps get found ft. Simon Thillay | Sub Club Live youtube.com/watch?v=0OhNiefG99g
Moeten de arbeidsgerechten en ondernemingsrechtbanken samensmelten tot één rechtbank? Welke rol kan AI spelen in de magistratuur? Aan de vooravond van haar pensioen deelt Van Den Bossche Gaby, eerste voorzitter van het Brusselse Arbeidshof, in de #Juristenkrant haar visie op enkele belangrijke uitdagingen: 🔷 Haar pleidooi voor een fusie van de arbeidsgerechten en ondernemingsrechtbanken 🔷 De nood aan meer samenhang in de steeds complexere en versnipperde sociale wetgeving 🔷 De kansen en risico’s van AI in de magistratuur Lees het volledige interview met Pol Deltour in de Juristenkrant, te raadplegen via InView Legal: lnkd.in/evKj7MK9
My DEF CON talk on hacking AI systems is now live. More research coming soon! 👇
🤖 Are you still entering quotations and orders manually? You may be wasting more time than you think. 🔽 📄 Customer requests still often arrive by email, PDF, Excel or DXF. You have to read. Analyse. Enter the data. Sometimes re-enter the same information. And all of this takes time that could be spent on higher-value tasks. 🖲️ That’s exactly where SynerAI Transaction comes in. Our AI-powered solution automatically analyses incoming documents, extracts the relevant data and transcribes it directly into Synerglass Symbiose as quotation or order lines. 🧑💻 Less manual data entry ❌ Fewer errors 🏃 Faster and more reliable processing And you always stay in control: the final validation remains in your hands. AI should not add complexity to your processes. It should remove repetitive tasks and help your teams work more efficiently every day. ⏱️ Want to learn more? Everything you should know is on our website! lnkd.in/euEVd-md Innovation. International growth. Synerglass-Soft #artificialintelligence #automation #glass #digital #transformation
Day one 🎓 with Nemesis takes you through setup, a tailored threat briefing, and your first live attack run in minutes. It probes, pivots, and reroutes just like an attacker would; giving you an answer to “what would actually happen if someone tried to breach your system?” After all, finding out about a major gap from an attacker on a Friday night is a nightmare. Discovering it on a Tuesday morning, with an AI that hands you the exact fix is a game changer. #Nemesis #ThreatDetection #securitycontrolvalidation
Our colleagues Eva Vandenberge and Timothy Robeers, PhD (SocScD) are enjoying a great day at the Data Analytics Conference in Utrecht, connecting with industry professionals, exchanging ideas, and exploring the latest developments in data, insights, and AI. One of the highlights will be Tim's roundtable session, which explores how three different AI-enabled methodological approaches can be applied to the same business question. Great to see our team meet up with everyone there including Wilko Rozema and Kjell Massen. #DAC2026#AI #wisdomthatcounts
Ne laissez pas Copilot dormir dans votre Microsoft 365 ! L’IA, c’est bien. Savoir l’utiliser concrètement dans son travail, c’est encore mieux ! Vous entendez parler de Microsoft Copilot, mais vous ne savez pas encore comment en tirer le meilleur parti ? Rejoignez notre formation « IA & Productivité avec Microsoft Copilot » et découvrez comment intégrer l’intelligence artificielle dans votre quotidien professionnel. Notre approche ? Du concret, de la pratique et des astuces directement applicables dans votre travail. Pas besoin d’être un expert en IA pour commencer ! Le 26 novembre à Liège, HUBCO vous propose une journée pour apprendre à utiliser Microsoft Copilot dans vos outils du quotidien et créer vos propres agents, sans coder. lnkd.in/eMXnUn_v Chez HUBCO, pas de formule magique compliquée. Juste des outils bien utilisés. Pas de bla-bla. Du résultat. #MicrosoftCopilot #IntelligenceArtificielle #Microsoft365 #Productivité #Formation #HUBCO
How much time does your organization spend looking for answers in its planning? Checking a forecast, tracing a shortage, finding the right figure... What if you could just ask IBP? Join our Supply Chain Masters Maïté Michielsen and Robin Van Acker on December 3rd for our webinar on 'AI in Integrated Business Planning', where we take you through 🟢What you can already do with AI in IBP today 🟢What talking to your IBP could look like 🟢What's in store for Supply Chain Planning and AI Register here: lnkd.in/eb6uWT_A #sap #autonomousenterprise #agenticai #supplychain #scm
Une question médicale pendant votre consultation ? Gagnez du temps dans vos recherches. 🔎 Avec l’Assistant aux connaissances médicales, intégré à CareConnect AI Assistant, posez votre question en langage naturel. Vous obtenez une réponse étayée, avec les références des sources utilisées, directement dans CareConnect GP. Moins de temps consacré aux recherches, plus d’attention pour votre patient. Vous restez maître de vos décisions, votre assistant IA vous accompagne. 👉 Essayez l’Assistant aux connaissances médicales et posez votre question dans CareConnect GP. hubs.la/Q04DqhL50 #CareConnect #IAenSanté #MédecineGénérale
AI may be changing the way we conduct research, but choosing the right methodology remains key to generating meaningful insights. Great to see our colleague Tim Robeers sparking some interesting discussions on this topic at #DAC2026 today!
Een medische vraag tijdens je consultatie? Je hoeft niet telkens verschillende richtlijnen te doorzoeken. 🔎 Met de Medische Kennis Assistent, onderdeel van de CareConnect AI Assistant, stel je je vraag in gewone taal. Je krijgt een onderbouwd antwoord met bronvermelding, rechtstreeks in CareConnect GP. Zo besteed je minder tijd aan opzoekwerk en hou je meer aandacht voor je patiënt. Jij houdt de regie, je AI Assistant denkt mee. 👉 Probeer de Medische Kennis Assistent en stel je vraag in CareConnect GP. Meer info: hubs.la/Q04Dqg2l0 #CareConnect #AIindezorg #Huisartsen
220,000+ claims transactions a month. 99.77% quality. 30% higher productivity. See how Sagility helped a workers' compensation organization streamline claims processing with AI-powered document automation: lnkd.in/eWahjAC4 #ClaimsProcessing #AutomatedDocumentProcessing
The engineers who know your process best rarely spend their day inside analytics software. The TrendMiner Agent in 2026.R3 is built for them. It works inside TrendMiner, understands the view you’re on and acts on it: it opens views, adds tags to your TrendHub view, changes the time range and filters context. Ask how to build a formula and you get an answer specific to your case, not a link to documentation. It runs on your LLM of choice, hosted by Microsoft Foundry or Anthropic, so your data stays within your company landscape. Want to see it on your own process data? Request a demo. More on 2026.R3: bit.ly/4eh9VgC #IndustrialAI #ProcessManufacturing #IndustrialAnalytics #AIAgents
See you in two weeks 🤩 Join us on 19 November at 11 AM CET for our webinar: Hermix AI Agents integrated into your own ChatGPT or Claude. We'll show you how to connect your AI tools to Hermix, use AI agents for market research and opportunity analysis, and prepare proposals with Hermix Studio. 🔴 Save your spot now: lnkd.in/eDDqmQVh
What will networks need to become in the AI Supercycle? Day 1 at the Brooklyn AI-Native 6G Summit explored how AI is reshaping the future of connectivity, and why AI, compute and networking are increasingly coming together. From lessons learned from the Software-Defined Networking era to AI-RAN, Physical AI, integrated sensing and live demonstrations of emerging technologies, the conversations offered a glimpse of what AI-native connectivity could look like. Catch the highlights from Day 1. #BrooklynAI6GS #6G #AI #AINative
Most AI conversations happen far away from the floor, so we moved ours to a dinner table 🍷 In Brussels, at The Hoxton, I sat down with JD Dillon and 15 leaders from retail and hospitality to talk about one question: where is AI actually paying off on the frontline? What stayed with me is simple: the best AI doesn't add another screen to a manager's day; it gives them time back to spend with their people. Amsterdam is next, on 10 November at NIM, and I'd love to see you at the table @khova #theshift
Banning shadow AI is the fastest way to stop hearing what your team needs. When people use AI tools you never approved, they’re telling you something. A task takes too long, or a report nobody reads still has to be written every week, and they found a shortcut without waiting for you. Our take: before you write an AI policy, find out what people already use AI for. That list is the most honest map of your process bottlenecks you’ll ever get. It shows where AI already saves time, and also where it carries real risk because company data ends up in tools nobody checked. Governance still matters. It just works better as a way to make useful habits safe than as a wall around the tools. So is shadow AI in your organization a risk to shut down, or a signal worth following up on? #AIStrategy #ShadowAI #AIGovernance #FutureOfWork #TheValueHub
🎉 Win a FREE 3-Day Ticket to ESPC26! Join Europe’s premier Microsoft 365 and AI conference and connect with experts, MVPs, and industry leaders. Gain insights on Microsoft 365, AI, Copilot and the wider Microsoft ecosystem, including Teams, SharePoint, Power Platform, Fabric, Entra and Azure. How to enter: 1️⃣ Open the form below 2️⃣ Submit your details 3️⃣ Wait for the winner announcement 🤞 📅 Closes: 6th Nov 2026 🔗 Enter: ESPC Sponsor Ticket Giveaway – Fill out form: lnkd.in/eHRafHad
An AI tool that suggests a drug candidate and one that drafts a deviation summary have something in common. You need to know what to trust them with. Across pharma, AI is helping teams screen molecules, recruit trial participants, detect manufacturing anomalies, and forecast demand. Generative AI is also making its way into research summaries, report drafts, and investigation support. But an impressive demo leaves plenty of questions unanswered. Where did the output come from? What happens when the model changes? Who checks the result before it influences a quality decision? Our latest article explores: ✔ Where AI supports the pharmaceutical lifecycle ✔ How generative AI is being used ✔ Challenges around data quality, validation, and oversight ✔ What to assess when comparing AI tools for GxP work The shortlist starts with the job you need AI to do and the consequences if it gets it wrong. 👉 Read the full article: lnkd.in/defFsUQn
AI Deepens the Mystery of Cellular "Vaults" For decades, biologists have been puzzled by "vaults," which are hollow, barrel-like structures found by the thousands inside cells. Their function remains a complete mystery. Now, Anthropic's scientists have used their AI model, Claude, as a computational partner to uncover new clues. Accelerating Scientific Discovery By analyzing a database of 15 billion proteins, Claude helped prove that bacteria and viruses also produce these odd structures. The AI guided the team's wet lab experiments, reducing protein purification processes from days to minutes and rapidly generating structural models. While the exact purpose of vaults is still unknown, this research demonstrates how AI can act as a powerful collaborator in biological discovery. lnkd.in/eFQdVav5
De arbeidsmarkt verandert voortdurend. Wat jarenlang een succesvolle aanpak was, levert daarom niet automatisch morgen nog resultaat op. Dat heeft Jobat zelf ook ervaren. Na de financiële crisis veranderde de manier waarop bedrijven kandidaten moesten bereiken. Van klassieke personeelsadvertenties en jobboards evolueerde Jobat naar employer branding, storytelling en steeds gerichtere campagnes. Door technologie, data en AI slim in te zetten, blijft Jobat zichzelf heruitvinden en inspelen op een arbeidsmarkt waarin de kandidaat steeds meer aan het stuur zit. Groei betekent niet altijd groter worden. Soms betekent het dat je jezelf durft heruitvinden. 🎙️In de nieuwste aflevering van Growth Sessions vertelt Werner van Jobat hoe innovatie en blijven evolueren de basis vormen voor duurzame groei.
Just ask me. Ask Spott is the AI assistant that brings everything your recruiting agency knows together in one place. • Which jobs fit the candidate I just spoke to? • Reach out to hiring contacts at companies similar to my best client. • Find jobs similar to my latest placement and shortlist the previous job's finalists. You ask. Spott does the work. You still make the call. Ask Spott, now live.
What if AI took a job nobody should have to do? That’s the side of the AI story we keep missing, and it’s where our next Off Script episode begins. Naina and Elizabeth sat down with Jaskaran Sandhu, COO and Partner at TMC, CEO of Open Pixel Systems and co-founder of XiniX AI. He builds AI so people no longer have to enter hazardous confined spaces. That idea took him all the way to a stage in Shanghai. All season we’ve asked whether big tech is ruining our lives. Jaskaran gave us a different angle: AI as the thing that protects. In the full episode we talk about: 🔹 What pulled him toward engineering, and when the people became as important as the technology 🔹 Why women belong in the room where AI systems get designed, and what goes wrong when they’re not there 🔹 Where the line sits between AI removing the danger and AI removing the livelihood 🔹 What building several companies at once really costs The trailer is out now. The full conversation drops Thursday 12 November. 🎧 Listen on Spotify: lnkd.in/ezkkYSTS 🎧 Listen on RSS: lnkd.in/ec8DCWrY Who do you think AI is really being built for? #AI #WomenInTech #Engineering #OffScript #Podcast
Get to know our Wizards: Meet Lambert🧙 In this video, Lambert Misselyn tells us a bit about his project with Mr. Energie. Lambert is actually our energy lover💚 Stay tuned for our last testimonial of the year next week👀 #data #energy #ai #project
Ask Spott is live. The AI assistant for your recruitment firm. It searches your entire database and acts on it: creates the job, matches candidates, drafts the outreach, moves the pipeline. And the more you use it, the smarter it gets. Your entire agency brain, one prompt away. Enjoy! spott.io/ask-spott
𝐖𝐞𝐥𝐤𝐞 𝐨𝐩𝐥𝐞𝐢𝐝𝐢𝐧𝐠𝐞𝐧 𝐫𝐚𝐚𝐝𝐭 𝐄𝐥𝐥𝐞𝐧 𝐃𝐞 𝐆𝐞𝐲𝐧𝐝𝐭 𝐚𝐚𝐧? Ellen De Geyndt is oprichter van Alfasent, een spin-off van Universiteit Gent. Vanuit haar passie voor taal en technologie ontwikkelde ze bij onderzoeksgroep LT3 de AI-modellen achter Alfasent. Met deze technologie helpt ze bedrijven om 𝗸𝗹𝗮𝗻𝘁𝗲𝗻𝗳𝗲𝗲𝗱𝗯𝗮𝗰𝗸 𝗮𝘂𝘁𝗼𝗺𝗮𝘁𝗶𝘀𝗰𝗵 𝘁𝗲 𝗮𝗻𝗮𝗹𝘆𝘀𝗲𝗿𝗲𝗻 𝗲𝗻 𝘃𝗲𝗿𝗯𝗲𝘁𝗲𝗿𝗽𝘂𝗻𝘁𝗲𝗻 𝘁𝗲 𝗶𝗱𝗲𝗻𝘁𝗶𝗳𝗶𝗰𝗲𝗿𝗲𝗻. Dit zijn de aanraders van Ellen: 𝟭. 𝗧𝗮𝗮𝗹𝘁𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝗶𝗲 𝗢𝗻𝘁𝗿𝗮𝗳𝗲𝗹𝗱 – LT3 - Language and Translation Technology Team Ik kan niet anders dan het boek van mijn ex-collega’s aanraden. Hoewel de technologie en mogelijkheden razendsnel evolueren, is dit boek ideaal voor wie echt wil begrijpen hoe tools zoals #ChatGPT eigenlijk werken. De complexe concepten worden in dit boek toegankelijk gemaakt voor het grote publiek. → lnkd.in/ejP2u6R6 𝟮. 𝗜𝗻𝘁𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗲 𝘁𝗼𝘁 (𝗴𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝗲𝘃𝗲) 𝗔𝗜 – Alfasent De AI Act vraagt dat organisaties ervoor zorgen dat hun medewerkers een basisniveau aan AI-geletterdheid behalen. De opleiding die we met Alfasent aanbieden is daar perfect voor. We geven alle medewerkers de nodige handvaten om AI op een bewuste manier in te zetten. → lnkd.in/eNj2bSaN 𝟯. 𝗘𝗗𝗨𝗯𝗼𝘅 𝗔𝗜: 𝗦𝗹𝗶𝗺𝗺𝗲 𝗺𝗮𝗰𝗵𝗶𝗻𝗲𝘀 𝗹𝗲𝗿𝗲𝗻 𝗯𝗲𝗴𝗿𝗶𝗷𝗽𝗲𝗻 – VRT NWS Ik maakte een tijdje geleden zelf kennis met één van de onderdelen van deze EDUbox. Leerkrachten kunnen er gebruik van maken om hun leerlingen een introductie te geven tot AI, zonder dat het te theoretisch wordt. → lnkd.in/enSkWVkc Vind je dit een toffe rubriek? Je kan hem ook in je inbox ontvangen, via onze driewekelijkse nieuwsbrief met opleidingstips. 👉 lnkd.in/erwy6eG7
Mistral just launched Timestral Just used it to predict hotel room pricing And it works EXTREMELY well. Data below: For context, Timestral is a time series prediction model - Claude and ChatGPT predict the next word. Timestral predicts the next numbers in a series - Timestral gives a forecast plus an 80% range it expects the real number to land in - Timestral is 3% the size of an LLM and can run on a CPU (not GPU) - Timestral forecasts any number over time, like prices or demand We tested it on two hotels in Portugal, one resort and one city hotel. - We forecasted the price per night, a year in advance - Each model is scored against a lazy guess that copies last week's prices forward RESORT HOTEL, A YEAR AHEAD For every €100 the lazy guess was off by: - Mistral Timestral was off by €17 - Amazon Chronos-2 by €18 - Google TimesFM by €20 - Meta Prophet by €21 - Microsoft LightGBM by €58 - ARIMA and ETS by €104 and €101, worse than the lazy guess BUT - each model gives an accuracy band... - Timestral gives a price and a band which covered 94% of real prices - Meta Prophet's band held 45% - Microsoft LightGBM's band held 24% But, on the city hotel, a year ahead, Timestral was off by €92 for every €100 the lazy guess missed - It still came first, with ARIMA next at €95 - Amazon Chronos-2 was off by €124, Meta Prophet by €138 and Microsoft LightGBM by €159 - City prices swing far less by season than the resort's, so a year out there is less pattern to find - Adding forecast max temperature brought Timestral down to €70 With the right tweaking and scenario, it's very possible to automate hotel price prediction (and we are in the process of doing it with a client) DM if you're interested.
The era of #AI is now. StoryChief - Content Marketing Platform #martech #campaign
A new capability for the FN SEA DEFNDER® at EURONAVAL 2026. Presented at Euronaval, FN DETECTOR brings AI-based detection to the FN SEA DEFNDER®. By analysing the camera’s video stream, it supports automatic target recognition and acquisition, helping the operator respond more efficiently to surface and aerial threats. Discover this latest evolution of our naval remote weapon station in the video. #AITracking #FNSEADEFNDER #FNDETECTOR #Euronaval2026 #NavalDefence
𝐁𝐥𝐚𝐜𝐤 𝐅𝐫𝐢𝐝𝐚𝐲 𝐢𝐬 𝐢𝐧 𝟑 𝐰𝐞𝐞𝐤𝐬. 🛒🎁 𝐈𝐬 𝐲𝐨𝐮𝐫 𝐜𝐮𝐬𝐭𝐨𝐦𝐞𝐫 𝐬𝐞𝐫𝐯𝐢𝐜𝐞 𝐫𝐞𝐚𝐝𝐲 𝐟𝐨𝐫 𝐢𝐭? For most retailers, peak season means one thing: an inbox that explodes. One customer care team we work with used to scale from 6 to 22 people every peak season, just to keep up. With AMOTEK Care, that same team now sees: 📉 62% faster reply times 👥 4.2 FTE saved per year 😊 +22 NPS after 6 months AMOTEK Care is our AI customer support for retailers: ✅ Fully resolves up to 68% of tickets, 24/7 ✅ Pulls order and delivery info straight from your ERP and TMS ✅ Live in a few weeks, on top of the systems you already use Curious what it could save you? Try the ROI calculator or book a demo 👇 🔗 link in comments #AMOTEK #CustomerService #Retail #AI #eCommerce Filip Smet - Matteo Martin - Shari Slagmolen - Christophe Dutoict - Niels Verdickt - Massimo P. - Senne Christiaens - Vincent Machiels - Mike Meyers - Nicolas Vanhuynegem - Matthias Coquin - Cindy Buyle - Luna Vanbiervliet
As a European citizen, I have mixed (good&bad) feelings about how USA and China influence what happens on our continent, so here goes one for the 'good': China is actively removing power concentration from the USA AI labs and distributing it over the world. I have no doubt that if China had been given the same access to GPUs as the USA AI labs, they would be dominating already. I imagine the relentless efforts at play to create this compute themselves. Jensen Huang is right: In the short term, distributing GPUs to China would hurt the USA AI labs (in terms of supply reduction/price hikes and adding competition), but in the long term, not distributing them will lead to the 'next' nvidia (ie compute hardware platform) to be Chinese. In the meantime, Europe is thinking about 'Green' and 'Safe' AI. I honestly don't understand where this comes from on the agenda. The 𝐍𝐞𝐭𝐡𝐞𝐫𝐥𝐚𝐧𝐝𝐬 has ASML creating the chip machines, 𝐁𝐞𝐥𝐠𝐢𝐮𝐦 has imec to make them produce chips and 𝐅𝐫𝐚𝐧𝐜𝐞 has abundant nuclear power (and water 😉 ) for the data centres. Where is this 'our diversity united leads to strength' thinking when we need it? Let's integrate, not regulate!
How do medical students practise difficult conversations when there are only so many actors to go around? That’s what our AI strategist Tom Wijnands talked about on the Microsoft stage at Trefdag Digitaal Vlaanderen Digitaal on 22 October. Trefdag is the yearly get-together for everyone working on digital government in Flanders, organised by Digitaal Vlaanderen. This year, more than 4,000 professionals came to Flanders Expo in Ghent, with sovereignty in a connected world as the main theme. Tom presented our case for KU Leuven: AI patient avatars. Practising with human actors works well, but it’s costly and scheduling the sessions can be challenging, so students get few chances to try it out. With the avatars, they can practise whenever it suits them. They talk to an AI patient who responds like a real person would, get feedback on their communication afterwards and can jump straight back in to have another go. Interested or want to know more? Check out our demo video and/or read the case study via the link in the comments.
6 numbers every CISO should know about AI agents 📊 82% of organisations found AI agents they didn't know about in the past year. Only 21% have a formal process to decommission them. And 92% of organisations with an AI-related breach lacked proper AI access controls. Agents act with real permissions. Most organisations can't see them all, let alone review them. Swipe through for all six 👉 More on 19 November in our live webinar with PlanB. GmbH. Link in the comments. #AIAgents #IdentitySecurity #CISO
🎬 Discover the AI-on-Demand Platform in action! 🚀 We've just released a new video walkthrough showing how you can explore, connect, and innovate with Europe's AI ecosystem. Here's what you can discover: 🧭 Navigate Europe's AI ecosystem Explore 1,000+ AI organisations across 40 European countries and 16 industries through interactive search and mapping. 🛒 Find the right AI solutions Browse the AI Marketplace for tools, computing resources, and services. Discover solutions and request demos in just a few clicks. 🛡️ Explore Sovereign AI Services Discover services designed to support data sovereignty, transparency, and compliance with the EU AI Act. 🔬 Access AI research and resources Explore 1.5M+ resources, including AI models, datasets, and learning materials to accelerate your innovation. 🌍 Become part of the European AI ecosystem Create a free account, increase your visibility, and connect with potential partners across Europe. Whether you're developing, adopting, or scaling AI, the AI-on-Demand Platform brings Europe's AI ecosystem together in one place. 🎥 Ready to explore? Watch the full walkthrough and get started today: lnkd.in/d55Eg6kw Discover the AI-on-Demand Platform: a new video walkthrough - AIoD aiodp.ai/news/discover-the-ai-on-demand-platform-a-new-video-walkthrough
Super excited to exchange ideas with the world's leading experts in media subtitling and localisation. For years, many assumed this part of media production would be among the first to be fully "automated" by AI. Yet the opposite became clear: while AI is 100% instrumental for ASR, editorial ownership, linguistic finesse and human touch are proving more important than ever. Huge congratulations to the organisers of Languages & The Media for bringing together such an outstanding community. Looking forward to two days of inspiring conversations, healthy debate and fresh perspectives!
Kan AI je codebase documenteren én die documentatie bruikbaar maken voor agents? Tijdens een ZAP-dag testte Koen E. EventCatalog op ons interne platform en werkte hij samen met Claude een workflow uit waarbij AI een project analyseert en automatisch een eerste documentatiebasis genereert. Wat werkte goed, waar liep hij tegenaan en wat maakt de MCP-koppeling interessant? Swipe voor de belangrijkste learnings. → Eens sparren? Get in touch met Fabio Van Dooren ☕️
Een collega bouwt een handige automatisering, een app die een manueel proces versnelt of een AI-agent die informatie uit verschillende systemen verzamelt en elke week enkele uren werk bespaart. Dat gebeurt gewoon binnen je eigen organisatie, iedereen is tevreden omdat het werkt. Tot die collega een nieuwe functie krijgt of beslist om ergens anders te gaan werken. Plots blijken dezelfde vragen op te duiken: Wie heeft toegang? Wie kan aanpassingen doen? Welke gegevens worden verwerkt? 👉 En vooral: wie neemt het over? Dan begint het zoeken: Hoe werkte dat proces precies weer? Wie beheert het vandaag? En kunnen we hier eigenlijk nog verder op bouwen? Soms wordt zelfs overwogen om een stap terug te zetten, gewoon omdat niemand meer weet hoe de oplossing beheerd of verder ontwikkeld moet worden. Op dat moment gaat het probleem zelden over technologie, het gaat om continuïteit. Want processen die jarenlang automatisch liepen, moeten plots weer manueel uitgevoerd worden. Dat is iets wat we bij 2Commit regelmatig tegenkomen bij organisaties die met AI, automatisering of het Power Platform aan de slag zijn gegaan. Niet omdat er iets fout gelopen is, integendeel, vaak zijn die oplossingen net succesvol geworden omdat ze snel waarde creëerden voor de organisatie. Alleen wordt vaak vergeten ‘Wat als die collega vertrekt?’ Daarom helpen we organisaties eerst overzicht creëren voordat er nieuwe oplossingen bijkomen. Niet met een lang consultancytraject, maar met een gerichte doorlichting die inzicht geeft in wat er vandaag draait, waar afhankelijkheden zitten en welke stappen logisch zijn om risico's te beperken. Want technologie die waarde creëert, mag niet afhankelijk zijn van één persoon. 💬 Heb jij al eens een stap terug moeten zetten? #2Commit #PowerPlatform #Governance #DigitalInnovation #KMO
What a brilliant narrative from Glenda Benson! Looking at where legal tech is heading today, the dominant conversation is almost entirely about AI models, semantic search, and automation. But Glenda’s story is a powerful reminder of a foundational truth: an important part of law is, and has always been, about people and relationships. When Afriwise was born, it didn't start with complex code or algorithms. It started with physical flip files of lawyers and laws, deep ground-level understanding, and, indeed, a drive to ensure that top-tier African legal expertise received the global recognition it deserved, rather than being reduced to a single line item on a pitch spreadsheet. The technology we build today, is only as good as the authentic domain expertise and human trust related to the data behind it. AI can process data at scale, but it cannot replace decades of earned relationships, local context, and real-world judgment. Proud to see how far your flip file has come, Glenda, and even prouder of the team and mission behind it. Worth a watch👇
Je was op amai!, en je hebt notities gemaakt van al jouw 𝘢𝘮𝘢𝘪!-momenten. En nu? Staan die notities in je gsm, of is er al iets veranderd in je bedrijf? Op tijd op de kar springen is één ding. Er ook effectief iets mee doen is wat het verschil maakt. Wat was jouw amai!-moment, en wat deed je er al mee? #amai #conversal #AIvoorkmo
As a fitness enthusiast, yesterday I gave the model I'm building (yes, also called Brubru, because it's built on the agents' know-how we already have) just one word: "gym". I wanted to see what it could do with it. It is still very primitive: robotic output and plenty of human-in-the-loop steering. But it reaches the finishing line. It framed its own hypotheses, crossed tables from a pool of 204 with EU survey and statistics data, and came back with some interesting new data. (Two of its hypotheses died on the way, which is the part I like!) EU data is a gift for this: primary-source, structured and ready to be fetched, if not always clean. The numbers come from read-only SQL rather than the model's memory, and they only count if they survive Spearman correlations, bootstrapped confidence intervals and a multiple-testing correction. It's just an experiment, a funny one! This model won't be public any time soon, because it's just for my own experimental purposes. Across Brubru I keep the temperature low, mostly between 0.1 and 0.5, because a legal reference is either right or wrong. And this experiment stops at correlations: I am not in the prediction layer yet! Brubru LLM is being trained on the europa.eu ecosystem: EuroVoc-informed tokenisation, a context window above 30,000 tokens, and QLoRA fine-tuning, because I don't have a supercomputer (just a Mac). Still in the making!
“Mostly right” doesn’t clear customs. AI can check every invoice, code and document. It shouldn’t sign the declaration. Under the UCC, whoever lodges it answers for it. So yes, trust AI with the groundwork. Keep the judgement calls with your people. Jana De Borger explains where that line sits in her article for LogisticsMatter. 🔗 Link in the comments
Discover what teams are actually creating with Drupal AI. If you think Drupal is behind the curve, its capabilities in the AI and agentic space will surprise you! This summit shows why.
Part 10 of my Obsidian search series is out: how AI search works on a personal vault. ➤ Keyword search misses synonyms. Semantic search blurs exact identifiers, and it never comes back empty. Hybrid search beats either one alone: in Anthropic's study, combining both and adding a reranker cut retrieval failures by 67%. ➤ The terminal agents I use (Claude Code, Codex, Gemini CLI, opencode) search with grep, not embeddings. So put the words you'll search with IN your files: titles, aliases, summaries. ➤ Your headings decide your chunks. ➤ The risks: stale indexes hide notes without telling you, and citations can be wrong. You also get a 20-query test to measure any search setup, scored with Recall@5 and MRR. lnkd.in/e7w9FYy5 All 12 parts in one book, with exercises: lnkd.in/eudNS4Tf The Obsidian Search Handbook | Find Any Note When You Need It (2026 edition) store.dsebastien.net/product/obsidian-search-handbook
🇪🇺 European or 🇺🇸 American AI models for clinical trial matching? Clearly, we're not the only ones asking the question! Great to see our latest research ranked #1 trending on Applied Clinical Trials online. 🚀 In "Comparing OpenAI and Mistral on Clinical Trial Matching", we explore how these models compare in terms of accuracy, data security, flexibility and cost, and what clinical research organisations should consider when choosing their AI models. A special thanks to Simeon Devos, our CTO, for leading the research, and to Andy Studna and the Applied Clinical Trials team for publishing it. 👉 Find the article link in the comments. What matters most when choosing AI for clinical research: performance, data sovereignty, cost or flexibility? We'd love to hear your perspective!
Stel: morgen start een nieuwe collega. Hij heeft geen manager en geen duidelijke functieomschrijving. Maar wel toegang tot je bedrijfsdata en systemen. No way dat een IT-team dat daarmee akkoord zou gaan. 😉 Toch krijgen AI-agents vandaag soms verrassend veel ruimte voor iemand vraagt: "ja maar, wie houdt hier eigenlijk toezicht op?" Want terwijl teams volop experimenteren, groeit ook de uitdaging: → Welke agents zijn actief en wie is er owner van? → Tot welke data hebben ze toegang? → Welke toepassingen glippen onder de radar als shadow AI? Op 12 november tonen Melissa Declercq en Wesley Droogenbroot tijdens ons webinar “Voorkom AI-chaos: krijg grip op je agents en AI-initiatieven” hoe je daar structuur in brengt, met aandacht voor AI Governance, security en concrete quick wins. 👉 Inschrijven kan nog: lnkd.in/dTTzxWRj #AIgovernance #AIagents #ITsAboutImpact
28DIGITAL is heading to Amsterdam, to be on the ground where innovation happens 🇳🇱 On Friday 13 November, the EIT AI Entrepreneurs Lab wraps up with its Grand Finale & Demo Day at AI House Amsterdam | Prosus, and 28DIGITAL will be in the room. Up to 15 founders, selected from more than 300 applicants, will pitch for up to five €10,000 prizes. This is the next generation of European AI & Robotics, and we want to meet them early. Representing 28DIGITAL: ⚫ Orestis Trasanidis ⚫ Parul Chaudhary ⚫ Rob Schubert Founders, investors, corporates and ecosystem builders: come and find us. The afternoon includes a keynote, the final pitches and networking, and it’s open to the ecosystem as part of the European AI Innovation Month. 📅 Friday 13 November, 14:00 – 17:00 📍 AI House Amsterdam, Gustav Mahlerplein 5 Places are limited. Register for free 👉 luma.com/5wextjm5
🎯 Deux entreprises peuvent utiliser les mêmes technologies sans avoir les mêmes besoins de formation. Leurs métiers, leurs outils, leur organisation et les compétences déjà présentes au sein des équipes influencent directement les apprentissages nécessaires. C'est pourquoi la formation continue gagne à être pensée au plus près des réalités professionnelles. 📚 Une entreprise peut chercher à accompagner l'intégration de l'IA dans ses processus. Une autre à renforcer les compétences en cybersécurité ou à préparer ses collaborateurs à de nouvelles responsabilités. L'enjeu est de proposer des formations qui tiennent compte des objectifs de l'entreprise, mais aussi du niveau et du parcours des travailleurs. La personnalisation des parcours permet ainsi de relier les besoins opérationnels au développement professionnel des collaborateurs. 💡 Comment identifier ces besoins et construire des formations adaptées ? Au Digital Day 2026, Atifa Boulafdal, de #DigitalcityBrussels, présentera les enjeux de la formation sur mesure et de la personnalisation des parcours. 🔗 Venez découvrir comment adapter la formation aux besoins de vos équipes, le 26 novembre à Bruxelles : lnkd.in/eFprpQ65 🎯 Twee bedrijven kunnen dezelfde technologieën gebruiken en toch heel andere opleidingsbehoeften hebben. De activiteiten, de gebruikte tools, de interne organisatie en de competenties die al binnen de teams aanwezig zijn, bepalen mee welke kennis en vaardigheden verder ontwikkeld moeten worden. Daarom is het belangrijk om bijscholing af te stemmen op de dagelijkse praktijk van bedrijven en hun medewerkers. 📚 Het ene bedrijf wil zijn medewerkers begeleiden bij de integratie van AI in de werkprocessen. Een ander wil de expertise op het vlak van cyberbeveiliging versterken of medewerkers voorbereiden op nieuwe verantwoordelijkheden. De uitdaging? Opleidingen aanbieden die aansluiten bij de doelstellingen van het bedrijf én rekening houden met het kennisniveau en het professionele traject van de medewerkers. Gepersonaliseerde opleidingstrajecten verbinden zo de concrete behoeften van bedrijven met de professionele ontwikkeling van hun medewerkers. 💡 Hoe brengt u die behoeften in kaart en ontwikkelt u opleidingen op maat? Tijdens Digital Day 2026 gaat Atifa Boulafdal van Digitalcity.brussels dieper in op opleidingen op maat en gepersonaliseerde leertrajecten. 🔗 Ontdek op 26 november in Brussel hoe u opleidingen beter kunt afstemmen op de behoeften van uw teams: lnkd.in/eFprpQ65 #FormationContinue #LearningAndDevelopment #DigitalDay2026 #DigitalcityBrussels Beci BeCentral hub.brussels SETCa Bruxelles Formation
AI-agents in jouw kantoor? Schrijf je nog snel in en ontdek hoe dat werkt in de praktijk. 👇 Een AI-agent controleert je dossiers voor je ernaar vraagt. Of bereidt je adviesgesprekken voor. Het verschil met automatisering? Automatisering volgt vaste regels, terwijl een agent complexe werkprocessen zelfstandig uitvoert. In deze workshop met Brecht Valcke van Oniroco leer je: - wat AI-agents zijn en welke taken je ze concreet toevertrouwt; - wat je vooraf regelt rond datakwaliteit, governance en privacy; - welke vaardigheden je team nodig heeft, met een preview van het actieplan. Er zijn nog enkele plaatsen vrij. Kies een locatie in je buurt en schrijf je in via lnkd.in/eVUYeivT - Brugge: dinsdag 17 november 2026 - Herenthout: maandag 23 november 2026 - Gent: donderdag 26 november 2026 Dit event telt voor 3 uur ITAA-accreditatie. ☑️
📞 Et si vos équipes commerciales échangeaient directement avec des prospects déjà qualifiés ? Avec un agent IA, la qualification de vos leads peut être automatisée. L’agent identifie et contacte vos leads entrants, échange naturellement avec eux, pose les bonnes questions et collecte les informations essentielles pour mieux comprendre leur besoin. Il peut également proposer et confirmer un rendez-vous, tout en intégrant les informations collectées dans votre CRM, pour permettre à vos équipes commerciales de prendre le relais avec toutes les informations nécessaires. ✅ 👉 Résultat : - des leads identifiés et qualifiés avant l’intervention de vos équipes ; - moins de temps mobilisé sur des demandes peu qualifiées ou des questions simples ; - des commerciaux mieux préparés, qui peuvent se concentrer sur les échanges à plus forte valeur ajoutée et la conversion. 🎥 Découvrez notre démo : un agent IA qui contacte un prospect après sa demande en ligne, qualifie son projet et organise un rendez-vous avec un conseiller. 💬 Et si vous intégriez un agent IA à votre processus commercial ? Parlons de votre projet 👉 lnkd.in/en-QcRfJ #IA #AgentIA #Automatisation #Innovation #Volubile
Waar is AI nuttig in je bedrijf? Deel 2: je projectdossiers. De meeste bedrijven werken met SharePoint of Google Drive. Bij elk nieuw project maak je een map aan, met submappen en documenten die ingevuld moeten worden. Meestal telkens hetzelfde. Dat kan vanzelf. Je stuurt een spraakberichtje naar je assistent: "Nieuw project voor dit bedrijf, start het op." En alles wordt in gang gezet. Tijdens het project volgt je AI-agent op. Hij vraagt documenten op bij de klant en zet ze op de juiste plek in je dossier. Na het project stuurt hij automatisch een mail om een review te vragen of het jaarlijkse onderhoud in te plannen. Het knip- en plakwerk valt weg. Dat kan zodra AI je bedrijf kent. Al mijn gratis gidsen en prompts staan op mijn resource hub. De link staat in de eerste comment. #AI #KMO #Automatisatie #CompanyBrain
Day 10 of 23: Agentic AI with Sigasi MCP Your coding agent can read HDL. It does not know your project. Through the MCP server built into Sigasi Visual HDL, a coding agent queries the same project model you work with in the editor, so what it generates is checked against the real design. Works with the frontier coding agents from Anthropic, OpenAI, Mistral, Cursor This is one of 23 short demo movies. Next will come tomorrow. Or watch them all right now: lnkd.in/eTmFFr8n Faster to better RTL is only 23 movies away #ASIC #FPGA #RTL #SigasiVisualHDL #MCP #AgenticEngineering #HDL
Deep vertical knowledge is required for AI impact across the firm. Jim Brown, Chief Operating Officer at Blick Rothenberg, on why they chose Ravical.
++ De volgende sprong in HR-tech ++ Stel je voor dat HR-software niet alleen helpt om administratie te organiseren, maar die administratie ook zelfstandig uitvoert! Met AI-agents die payroll voorbereiden, verlofsaldi opvolgen, HR-processen bewaken en zelfs nieuwe softwarefunctionaliteiten bouwen, komt een nieuwe generatie HR-technologie in zicht. >> BEKIJK DE AFLEVERING: lnkd.in/evzk3z7Q >> LUISTER via SPOTIFY: lnkd.in/eA8XCzDD In deze aflevering van #ZigZagHR Brainpickings gaat Lesley Arens in gesprek met Michiel Bearelle, medeoprichter van de Gentse HR-techstart-up Peepel en eerder betrokken bij de oprichting van Officient. Na jaren van digitalisering staat HR volgens Michiel voor een volgende grote sprong: van software waarmee we werken naar software waaraan we werk kunnen uitbesteden. ✅ Wat betekent dat concreet voor HR-professionals en kmo's? ✅ Hoeveel autonomie geef je aan AI-agents? ✅ Waarom verandert dat de manier waarop we HR-software selecteren? ✅ En wat betekent deze evolutie voor sociale secretariaten en HR-dienstverleners? We bespreken ook hoe HR-professionals zelf toepassingen kunnen laten bouwen door eenvoudig te beschrijven wat ze nodig hebben, zonder zelf te programmeren. Boeiend! Boeiend! Boeiend! Merci Michiel voor het bijzonder boeiend gesprek!
𝗝𝗲 𝗯𝗲𝗱𝗿𝗶𝗷𝗳𝘀𝗱𝗮𝘁𝗮 𝘇𝗼𝗺𝗮𝗮𝗿 𝗼𝗽𝗲𝗻 𝗲𝗻 𝗯𝗹𝗼𝗼𝘁 𝗼𝗽 𝗱𝗲 𝗠𝗲𝗶𝗿 𝗼𝗳 𝗱𝗲 𝗚𝗿𝗼𝘁𝗲 𝗠𝗮𝗿𝗸𝘁 𝗹𝗮𝘁𝗲𝗻 𝗹𝗶𝗴𝗴𝗲𝗻? 🤯 Waarschijnlijk niet. En toch gebeurt dat elke dag wanneer medewerkers zonder duidelijke afspraken gebruikmaken van openbare AI tools. AI moet je bedrijf sneller en slimmer maken. Niet kwetsbaarder. De sleutel ligt in een sterke basis: duidelijke richtlijnen, een private AI omgeving achter je eigen firewall en medewerkers die weten welke informatie ze wel en niet mogen delen met een AI model. Geen ingewikkelde paranoia. Gewoon gezond verstand, gecombineerd met de juiste beveiliging op de juiste plek. Want hoe krachtig AI ook is, vertrouwen blijft de fundering waarop alles gebouwd wordt. En dat vertrouwen begint bij een veilige omgang met data. Benieuwd waar jouw organisatie vandaag staat op het vlak van dataveiligheid en AI volwassenheid? 𝑩𝒓𝒆𝒏𝒈 𝒉𝒆𝒕 𝒉𝒂𝒂𝒓𝒔𝒄𝒉𝒆𝒓𝒑 𝒊𝒏 𝒌𝒂𝒂𝒓𝒕 𝒎𝒆𝒕 𝒐𝒏𝒛𝒆 𝒈𝒓𝒂𝒕𝒊𝒔 𝑨𝑰 𝑴𝒂𝒕𝒖𝒓𝒊𝒕𝒆𝒊𝒕𝒔𝒔𝒄𝒂𝒏. 👉 lnkd.in/eSb3WzJH #DataGovernance #AI #KMOBelgië #DigitaleTransformatie #Geniki
We hebben meer informatie dan ooit. Maar vinden we ook sneller wat we nodig hebben? 💡 Rapporten, nota’s, presentaties, beleidsdocumenten... De hoeveelheid kennis binnen organisaties blijft groeien. En dus groeien vaak ook de mappenstructuren, metadata en regels mee. Maar kennisbeheer hoeft niet complexer te worden. Met AI-autoclassificatie kunnen documenten automatisch geanalyseerd, samengevat en voorzien worden van relevante metadata binnen de kennisstructuur van je organisatie. Een inhoudelijke expert behoudt daarbij (meestal) de finale controle via een Human in the Loop-aanpak. Zo gaat het niet alleen om documenten sneller verwerken, maar vooral om kennis beter vindbaar, consistenter en bruikbaarder te maken. 👉 In onze nieuwste blog lees je hoe AI-autoclassificatie documentbeheer kan laten evolueren naar écht kennisbeheer: lnkd.in/emAgaTPj #AI #Kennisbeheer #DocumentManagement #WorkplaceInnovation
There is a version of hotel AI that will tell you RevPAR is up while reading a table it only half understands, and the dangerous part has nothing to do with the model being poor. It is dangerous because a wrong answer arrives in exactly the same tone as a right one. A revenue manager who has lived with these numbers for fifteen years will catch it. A duty manager asking about Saturday's arrivals will not, and should not have to. Which makes the questions worth putting to any vendor fairly unglamorous ones. Where did this answer come from, and can you click through to the rows behind it. What does the system do when it does not know, because an honest "I cannot answer that from your data" is worth considerably more than a plausible guess. And where does the data live, including whether your commercial performance quietly trains a model that other hotels then benefit from. We would rather be checkable than impressive, so for the record: Kassandra AI answers from your connected sources with the working shown and your data never trains shared models. #Juyo #HotelTech #AI #ResponsibleAI #Hospitality #DataSecurity
You've given employees access to AI. But have you shown them what to do with it? Without clear guidelines, practical examples and somewhere to ask questions, even the best AI tools can go unused. Your intranet can help employees find what they need to get started and feel confident using AI in their daily work. 👉 More tips in the full article. Link in the first comment. #AIAdoption #Intranet #InternalCommunication
What’s next in Life Sciences & Healthcare? AI is moving fast. The real opportunity is turning it into meaningful impact. On Thursday, 26th November, Spire, Medminds and De Cronos Groep bring together professionals from Life Sciences, Healthcare and technology for Next in Life Sciences & Healthcare at LOGIVILLE in Niel. An evening focused on how AI, data, Salesforce and connected technology are already changing the way organisations operate, make decisions and create better experiences. You can expect: - An exclusive AI in Operations tour at LogiVille - Expert perspectives on AI, data and technology - Concrete use cases and live demos - Food, drinks and meaningful conversations with peers 📅 Date: Thursday, 26 November 2026 🕔 Time: 17:00 – 21:30 📍 Location: LogiVille, Niel Register via the link in the comments. Spire × Medminds × Cronos #Medminds #LifeSciences #Healthcare #AI #Salesforce #DigitalTransformation #MemberOfSpire
My last post on the SF Tech Week - Day 3 end Forward Deploying Engineer (FDE) is something that becomes increasingly popular in the AI field. It’s the fact to use agents to further integrate products at a customer and it just so happen we do that since quite some time at Deuse and we do it well. Special thanks to our minister Pierre-Yves Jeholet, Digital Wallonia and Antoine Hublet to make this happen in Belgium with the Start IA and Tremplin IA 🚀 . Some of the things we talked about are: - Getting the right data is very hard for customers. - Think hard about when to put the human in the loop. - Teach the agent what it can do. Observability and traceability matter. - Key to make it work: trusted context (skills, connectors, policies). And one of my takeaways, we are entering an era where the line between custom development company such as Deuse and SaaS company is getting blurred. If you want an agent / tool integrated in your company or you are wondering how it works, feel free to reach out ! We’ll be happy to help. I also had interesting talks there with Roger Brockett, Calvin 🦙 Sedao, Ravindra Neriyanuri, Shouvik Ghosh, David Van Reyk, Alora Frederick, Rachana Mansinghka, Hank B.
3 jaar geleden stond LinkedIn vol over ABM. Ik deed zelf volop mee. Daarna werd het stil. Maar ABM is nooit weggeweest. Het blijft de manier om je droomaccounts te bereiken, online én offline. Het probleem was nooit de strategie. Het was de tijd. Met agentic AI kan je vandaag een ABM-platform bouwen dat de bestaande tools op hun eigen terrein verslaat. Dus hebben we dat gedaan. Maak kennis met Super ABM. Onze eerste beta-klanten zitten erin: 📉 Workload per account: van 45% naar 15% 📉 Researchtijd: van 55% naar 5% 📉 Inbound T1 ICP Fit +300% Elke salesprofessional die ik het toon, reageert hetzelfde. Ze zijn verbaasd hoeveel tijd ze terugwinnen en hoe makkelijk opvolgen wordt. En eindelijk komen die middelgrote en grote accounts zelf warm binnen, in plaats van dat je ze eindeloos achterna loopt. → Elk target account en elke DMU wordt opgevolgd, bijna op autopilot → Sales blijft aan het stuur → Een platform plus een Chrome-plugin, gewoon waar je al werkt We laten niet iedereen tegelijk binnen. Er zijn nog 5 plaatsen voor organisaties met minstens 250 target accounts en hubspot. Die demo geef ik persoonlijk. superabm.com
Welke skills vraagt de fabriek van morgen? En wie leert ze aan? 🏭 AI verandert de jobs in de maakindustrie, zowel op de werkvloer als in productontwikkeling. Agoria stelt daarom de nieuwe studie AI Skill Shift in Manufacturing voor tijdens een roadshow. Welke rollen en taken veranderen door AI en digitalisering? Welke competenties worden belangrijker? En hoe kunnen bedrijven en opleidingspartners ervoor zorgen dat leerprogramma's blijven aansluiten bij wat de werkvloer nodig heeft? 📅 Dinsdag 1 december, 14u30 tot 18u 📍 House of Manufacturing, Kortrijk (of op 26 november in T2-campus, Genk) 👉 lnkd.in/ewH7kU22 Vickie Dekocker (PhD) Veerle Vermeulen Julie Ampen #AISkillShift #Manufacturing #FutureWorkforce #SkillsDevelopment
As AI becomes part of SAP Engagement Cloud, the decisions that shape its value sit less in the software and more in the organisation around it. Who is accountable for content that AI helps create? How much room does a team get to build new skills? Which teams share responsibility for the data the AI works with? And how far may AI act before a person steps in? These are choices for marketing leadership, and they can be made before the capabilities are in daily use. We set out four points to consider.
Een klant zegt: “Ons merk moet sterker worden.” Maar wat betekent dat precies? Een scherpere positionering? Meer inzicht in klanten? Een onderscheidender verhaal? Of meer consistentie in de communicatie? Merkadviseurs krijgen regelmatig met deze vragen te maken. AI wordt steeds vaker ingezet om het onderzoekswerk en de strategische voorbereiding te ondersteunen. Daaruit zijn 7 praktische AI-projecten geselecteerd: 1. Merkkompas: verduidelijk identiteit, positionering en merkbelofte. 2. Concurrentiebeeld: analyseer concurrenten en marktkansen. 3. Doelgroepinzicht: structureer klantbehoeften en aankoopmotieven. 4. Merkstem: ontwikkel consistente communicatierichtlijnen. 5. Merktoets: beoordeel boodschappen aan de hand van de merkstrategie. 6. Merkverhaal: bouw een samenhangend en geloofwaardig merkverhaal. 7. Merkplan: vertaal strategische keuzes naar concrete prioriteiten. Het doel is niet om creativiteit of expertise te vervangen, maar om terugkerend werk te verminderen en meer ruimte te creëren voor strategisch denken. Alle 7 projecten worden op aanvraag samen aangeboden als één compleet downloadbaar pakket met afzonderlijke, volledige broncodes. Ze zijn zelfstandig te gebruiken, aan te passen en uit te breiden in geschikte AI-chatbots, zonder programmeerkennis, technische expertise of complexe installatie. Wil je het volledige pakket ontvangen? Stuur een DM.
De beste manier om AI te gebruiken? Je eigen denkvermogen versterken, niet vervangen. Steeds meer schrijvers en contentmakers ontdekken hoe AI hun dagelijkse werk kan ondersteunen. Niet alleen door teksten te genereren, maar vooral door ideeën te verkennen, onderzoek te structureren en eigen werk kritisch te beoordelen. Internationaal onderzoek laat zien dat brainstormen, informatie verzamelen en teksten verbeteren belangrijke toepassingen zijn. Dat biedt interessante mogelijkheden voor iedereen die schrijft, publiceert of anderen leert schrijven. Op basis van deze toepassingen heb ik 6 praktische AI-projecten geselecteerd: 1. Tekstverbeteraar Helpt je eigen teksten duidelijker, sterker en beter leesbaar te maken, zonder je persoonlijke schrijfstijl te verliezen. 2. Ideeverkenner Stelt gerichte vragen, onderzoekt verschillende invalshoeken en helpt je eigen ideeën verder te ontwikkelen. 3. Bronnenwijzer Brengt structuur aan in onderzoeksmateriaal en helpt je ontdekken welke informatie of onderbouwing nog ontbreekt. 4. Argumententoets Onderzoekt je redeneringen, stelt kritische vragen en maakt zwakke plekken in je argumentatie zichtbaar. 5. Inhoudsplanner Helpt je onderwerpen, ideeën en publicatiemomenten organiseren in een overzichtelijke en haalbare planning. 6. Schrijfbegeleider Ondersteunt je schrijfontwikkeling met oefeningen, persoonlijke feedback en gerichte verbeterpunten. Het bijzondere aan deze projecten is hun uitgangspunt: AI ondersteunt het proces, maar jij blijft zelf denken, kiezen en schrijven. Elk project is bedoeld voor herhaald gebruik en kan helpen tijd te besparen, efficiënter te werken en meer overzicht te creëren. Je hebt geen programmeerkennis, technische ervaring of ingewikkelde installaties nodig. Alle 6 projecten zijn samen verkrijgbaar als één compleet, downloadbaar pakket met afzonderlijke, volledige broncodes. De instructies kun je zelfstandig gebruiken in geschikte AI-chatbots. Je kunt ze bovendien aanpassen, uitbreiden en afstemmen op je eigen werkzaamheden. Zo wordt AI geen vervanging van je creativiteit, maar een praktisch hulpmiddel om die beter te benutten. Wil je het volledige pakket ontvangen? Stuur me een DM. Het complete pakket is verkrijgbaar op aanvraag.
Live from 6sense breakthrough....My latest Blog Post on LLMs. How much does the chatbot really decide? linkedin.com/pulse/how-much-does-chatbot-really-decide-koen-de-witte-marle?trackingId=7w3mR0TnqHOVC2nLIm7Y2Q%3D%3D
Here's our AI assistant explaining what you can do with your assistant...
From DX to AX: developer experience → agent experience. For years, great API docs meant one thing: making developers productive fast. That’s still true. But today, part of every integration gets written by a coding agent, and most docs weren’t built for that reader. So we rebuilt the Apideck developer docs for both. For developers: → Start from what you’re building instead of from an endpoint list. 37 use cases, each opening a reference filtered to exactly the operations you need. For accounts payable, that cuts 186 operations down to 36. → Connector gotchas before you commit. QuickBooks alone has 75 documented quirks, so you don’t discover them in production. → Faster pages, dark mode everywhere, and a reference that works on your phone. For agents: → Add .md to any docs URL and get clean Markdown back → llms.txt telling agents what to do first → A Docs MCP server, no API key needed. One command in Claude Code and your agent reads the same docs you do. The part I’m most proud of: we didn’t just guess what agents need. We gave coding agents real tasks against the docs and read every trace. They caught things no human review did. One operation’s Markdown dropped the auth it inherited, so the agent went hunting across other pages before it could make a call. Now auth travels with every operation. If your docs only work in a browser, half your readers can’t use them anymore. Video walkthrough + full write-up in the comments 👇
The Juyo community is growing and we are seeing some very interesting AI use cases developing. If you are part of the community this is a great session to join✌️
📣Day 1 of SynaptiQ Connect Days 2026: the rules of Asset Performance Management have changed. Have you? 💪Every year, SynaptiQ Connect Days brings together customers from across the world, creating more than just an event: a true community of SynaptiQ users. A place to connect, exchange ideas, share experiences and challenges, and, above all, learn from one another. Four key pillars are shaping this year's conversations: 💠BESS 💠Revenue 💠AI 💠Scale 🗣️Today, we had the pleasure of hearing inspiring real-life case studies from Mexens ex.Technique Solaire, BNZ, PerPetum Energy, GS Power Partners and EDP Renewables, who took the stage to share their experiences and insights. 🔋The most rewarding part of the day? Seeing people connect, build relationships and realise that they are all facing similar challenges and shaping the future of renewable energy together. And this is what SynaptiQ Connect Days is all about. 🌍 💪A big thanks to all the participants! Esther García Cabrero Thomas DHUICQ Mathieu Tahon Taylor Oldro Margarida Correia Pires #SynaptiQConnectDays2026 #SynaptiQ #AssetPerformanceManagement #RenewableEnergy #BESS #PV #AI
Gisteren bracht Brightlands Data & AI Campus 25 ondernemers uit de leisure- en hospitalitysector samen voor de AI Leisure & Hospitality Kickstart, om de mogelijkheden van AI te verkennen. Met AI-D deelden we concrete praktijkcases die we binnen de sector hebben geïmplementeerd, met een bijzondere focus op Conversational AI. Een mooie gelegenheid om te tonen wat vandaag al mogelijk is en hoe AI kan bijdragen aan efficiëntere processen en een betere gastenbeleving. Op de hoogte blijven van onze AI-projecten, events en inzichten? Schrijf je in op onze nieuwsbrief: lnkd.in/en5aCMC5
“I don’t want to talk to your agents in your tool. I want my agent to talk to your tool.” That quote perfectly sums up how our way of working has changed over the past few months. Today, we work almost entirely from Claude: - Keyword research using the DataForSEO and SEO Gets MCPs. - Technical SEO by controlling Screaming Frog via MCP, or running it through CLI on our VPS with a Claude managed agent. - Data analysis using SEO Gets, Gscwizard.com, Rankshift, and our own tracking tools. - Content creation connected to AirOps or our own knowledge base, combined with tools like Firecrawl and Parallel Web Systems. - Time tracking through an unofficial Teamleader MCP. Everyone on our team has built their own workflows, connected their preferred MCPs, and set things up in the tools they like working with. Yet I still see so many companies building their own AI agents, adding workflows inside their platforms, and trying to become the next all-in-one solution. We're moving in the opposite direction. That's why we built Depends OS. Not another AI agent, but one central place where everything comes together: - All client brand information, context, and relevant sources. - Playbooks and skills that can be customized per client and content type. - Connected tools and data sources, so clients don't need their own DataForSEO account or have to configure a GSC MCP. - Content that anyone on the team can review, edit, and manage, including through MCP. But here's the important part: Almost nobody has to work inside Depends OS itself. Everyone can connect to Depends OS via MCP, directly from their own Claude environment. They can brainstorm with it, retrieve client context, combine data from other MCPs, create content, or edit existing articles. All without leaving the environment they prefer. And I think that's exactly where the challenge lies for many software companies. You can never build features and workflows fast enough to keep up with how creative your users are. But you can give them the right context, tools, and capabilities to build their own ideal workflows. Instead of building another tool with an AI agent inside, build a tool that any AI agent can work with. Curious how others are approaching this. Are you still working mainly inside individual tools, or are you increasingly using your own AI agent as the central interface?
Tuesday, 10am: the best time to send your newsletter? For Léa, who reads her emails on the train at 7am, it’s too late. For Sophie, who opens them on the sofa in the evening, it’s far too early. The best send time depends on each customer. And there are as many of them as there are profiles in your database. In Actito, AI Optimisation sends every email at the moment each profile is most likely to open it. Tick a box in your campaigns, switch it on in your scenarios. We’ll show you where to find it, plus 3 tips to get the most out of it 👉 What time do you send at the moment? #Actito #EmailMarketing
Almost ready for #SIRM2026! We’re at the Allianz MiCo Center in #Milan, putting the finishing touches on the AGFA HealthCare booth B58 - getting ready to welcome the Italian #radiology community. Tomorrow, the doors open! Visit us to explore our latest #EnterpriseImaging and #RIS innovations, including Streaming Client, RIS Elefante® Web, RUBEE® #AI, RUBEE® Orchestrator and Enterprise Imaging #Cloud. See our solutions in action through live demos, connect with our #Italy team and join our educational sessions throughout SIRM. Don't miss out our booth education sessions! Tomorrow, 5 November at 15:30, see "RIS Elefante & Structured Reports" with AGFA's Roberto Giorgi. 👉 Book a demo or meeting with our team and discover how we can help you stay in flow. okt.to/yo0Tlr And while you’re there, pick up a little something to help you stay in flow beyond SIRM too 😀 Our team is ready: Davide Fumagalli, Andrea Polticchia, Marco Arcelloni and Renato Iaccarino. See you tomorrow! #EmpowerImaging
You assume a missed call leaves a voicemail. Most callers don't leave one. In a CallRail survey of 1,000 US consumers last year, 82% said they'll call a competitor if you don't answer. So I set up a made-up spa with the AI receptionist we built at IB Digital answering, and called it as a customer. It picked up, checked the calendar and booked the massage. Mid-call, I asked for a birthday discount. What it said is in the video, along with the whole call, uncut. Tell me what kind of business you run, and I'll pick one, set up a made-up version and call it in a future video.
The era of #AI is now! #thebillablehour
Mardi 10h, le meilleur moment pour envoyer votre newsletter ? Pour Léa, qui lit ses emails dans le train à 7h, c’est trop tard. Pour Sophie, qui les ouvre le soir sur son canapé, c’est bien trop tôt. Le meilleur moment d’envoi, c’est celui de vos clients. Et il y en a autant que de profils dans votre base. Dans Actito, l’Optimisation IA envoie chaque email au moment où chaque profil a le plus de chances de l’ouvrir. Une case à cocher dans vos campagnes, une option dans vos scénarios. On vous montre où la trouver, et les 3 astuces pour en tirer le meilleur 👉 Et vous, vous envoyez à quelle heure aujourd’hui ? #Actito #EmailMarketing
AI gebruiken in je bedrijf. Klinkt simpel. Tot je moet beslissen waar je begint. 👀 Want hoe gaan bedrijven vandaag écht om met AI? Waar zien ze kansen? Waar botsen ze? En wanneer wordt experimenteren ook effectief implementeren? Stijn deelt zijn kijk in onze nieuwste podcast. 🎙️ Binnenkort meer. 🤖
🤖 AI is no longer a tech requirement. Marketing and growth roles asking for an AI skill jumped from 10% to 13% of remote postings in a single month, and administration and executive support went from 4.1% to 6.2%. In September, 58.8% of remote postings on Jobgether that asked for an AI skill, and had a known role family, sat outside data and software engineering. Many experienced professionals already use these tools every day and never mention them, so screening systems read their profile as missing a skill they actually have. Name the specific AI tools you use in your function, on both your CV and your LinkedIn profile, and tie each one to a result. 👉 AI CV Review and AI LinkedIn Optimizer show what your profile is missing for the roles you’re targeting, and the full breakdown by role family is here lnkd.in/gvBNacJ4
London calling! 🇬🇧 Next Tuesday we’re hosting “Beyond the Booking” at our customer Clays. An evening for venue leaders on what AI is changing in how guests find you, enquire and book. Running a venue and curious where things are heading? A few seats left: luma.com/lu1acp2g Then on Wednesday and Thursday you’ll find us at Event Tech Live(stand KS1), showing what the next generation of venue management looks like. Let’s grab a coffee if you’re around.
The most useful Copilot feature we’ve seen so far? Not AI agents. Not fancy demos. 📊 Formula suggestions in Excel. Type '=' and Copilot can suggest a formula based on your data. Simple. Fast. Useful. And that’s the point. The Copilot features that stick are often the ones that remove small frustrations from everyday work. #MicrosoftCopilot #Microsoft365 #Excel #AI #Productivity #Semasu
“Agents in Practice: A Governed Approach to Agentic AI in Clinical Oversight” is now available on demand. Agentic AI has the potential to help clinical teams move more efficiently from data and insights to decisions and actions. But real-world adoption depends on how these capabilities are governed, traced and integrated into clinical workflows. In this session, Usama D., Chief Product and Technology Officer; Richard Young, Chief Strategy Officer; and Richard Davies, Principal Solutions Consultant, share CluePoints’ approach, including: → A demonstration of CluePoints’ current agentic capabilities → How agentic AI can support clinical data review and oversight → A preview of where those capabilities are headed → How CluePoints is approaching human oversight, traceability and governance Watch on demand: hubs.ly/Q04Dghn10 #AgenticAI #ClinicalTrials #RBQM #ClinicalOperations #ClinicalDataManagement
Bad data quality ruins everything. A model’s accuracy score doesn’t tell you whether its production data is fit for use. A model can perform well in the lab and still encounter incomplete, inconsistent, or incorrect data in production. And it may keep running while those inputs undermine its results. Data quality checks and monitoring help teams spot these problems. But an alert doesn’t make the affected data fit for use. Someone still needs to understand what went wrong, decide what should change, and verify that the correction worked. That could mean correcting records, fixing a transformation, or addressing the source of the problem. Until that work is done, the issue you detected can keep affecting downstream results. That’s why finding, understanding, and fixing data quality issues need to be connected. AI-ready data depends on teams being able to act on what they find. And finding a data issue should lead to a resolution your team can verify. On November 17, we’ll show how Soda connects finding, understanding, and fixing data quality issues. Join us for "Data Quality is Solved", streamed live from New York: lnkd.in/e5NGYG3z
Is AI really "more profound than fire or electricity"? In our latest newsletter edition, our CTO, Carl D'Halluin, contextualizes AI against other transformational tech trends like virtualization and cloud computing. The main takeaway? "The challenge is to understand the data estate well enough to decide what information AI should use," and do so well enough to help ensure the hype eventually matches reality. Read the full issue below! What Previous Tech Revolutions Tell Us About The AI Data Problem linkedin.com/pulse/what-previous-tech-revolutions-tell-us-ai-data-problem-datadobi-afjje?trackingId=sjJd3J6aa638ey8yjHX8VA%3D%3D
😎 Some time ago, the Teleport team brought us along to New York for a game day where our only job was to build the thing the attendees had to break. Using Nemesis, we engineered the environment, the clues, and the automated backend that kept the CTF running. Looking back at it now, seeing our adversary emulation platform power the chaos from behind the scenes was fun. Link: lnkd.in/ebM7ekUJ #NYC #Nemesis #AIsecurity #ThreatToTest Running Teleport's CTF Game-Day with Nemesis | PSI persistent-security.net/post/teleport-ctf-game-day-nemesis
What if you could differentiate your lessons without doubling your workload? With the BookWidgets AI Assistant, you can! ✨ In this video, we show you how to adapt an existing quiz to better meet your students’ needs. For example, you can split questions into pages with different difficulty levels, add more challenging open-ended questions, and include a self-reflection rubric. Review the suggestions, request changes, and approve the result. You guide the process every step of the way! 🎥 Want to discover everything the BookWidgets AI Assistant can do? Visit our website to learn more about the BookWidgets AI Assistant and register for one of our launch webinars! 👉 lnkd.in/eMXJSbw7 #BookWidgets #AIinEducation #EdTech #TeachingIdeas #LessonPlanning #InteractiveLearning #TeacherLife #DigitalLearning #AIforTeachers #TeacherInspiration
We're looking for a Head of Growth 🚀🍀✨ Want to close deals and help shape the future of a fast-growing AI company? 👀 Interested? More info in the comments. Ready to apply? Send your CV to info@amotek.be #hiring #sales #AI #growth #headofgrowth #B2B #Antwerp Filip Smet - Massimo P. - Matteo Martin - Luna Vanbiervliet - Shari Slagmolen - Christophe Dutoict - Vincent Machiels - Niels Verdickt - Senne Christiaens - Cindy Buyle - Matthias Coquin
🏥 PME HealthTech : et si vous présentiez votre solution d'IA ou de robotique directement à un réseau d'hôpitaux européens ? 🚀 TEF-Health vous offre cette opportunité unique, gratuitement ! Vous développez une solution innovante dans le domaine médical, basée sur l'intelligence artificielle et/ou la robotique ? Grâce à TEF-Health, vous pouvez présenter votre solution lors d'une session de pitch en ligne de 15 minutes, devant des représentants d'un réseau européen réunissant 10 hôpitaux et un institut de recherche dans 8 pays. 🎯 L'objectif ? - Présenter votre innovation et les défis que vous rencontrez. - Recueillir les premiers retours d'experts hospitaliers. - Identifier des opportunités de collaboration, de test et de validation en environnement clinique réel. - Accélérer votre parcours vers le marché européen. 📅 Prochaines sessions : 🔴 6 novembre 2026 : COMPLET ! 🟢 4 décembre 2026 : inscriptions ouvertes, places limitées ! 💡 TEF-Health, c'est quoi ? Une initiative européenne de Testing and Experimentation Facilities qui accompagne les entreprises dans le développement, le test et la validation de solutions d'IA et de robotique appliquées à la santé. Cofinancée par l'Union européenne et, en Belgique, par la Wallonie, cette initiative dispose d'un nœud belge piloté par Multitel, en collaboration avec ses partenaires. 👉 Ne manquez pas cette opportunité de confronter votre innovation aux besoins réels des hôpitaux européens ! Informations et inscriptions : lnkd.in/e4Rjp8RU HUmani - CHU Charleroi-Chimay CETIC SPW Economie Agence du Numérique (AdN) BioWin IGNITY Anne-Laure CADJI Bridoux Alexandre Quenon, PhD François Narbonneau Valérie HERTVELDT Christophe Dugast, PhD Karim Zouaoui Boudjeltia Gauthier Lejeune Jean-Christophe Deprez Caroline de Clock Chloé Hanin Thomas Louis Fabrizio Giannotta #TEFHealth #HealthTech #MedTech #IntelligenceArtificielle #RobotiqueMedicale #Innovation #PME #Wallonie #DigitalHealth
Will an AI agent take your job? Ask Ferris! We gave her an AI colleague and her first reaction is the face you see at the start of this video. But the agent doesn't do her job. It does the digging and the prep for the reports, day and night, so Ferris gets the part that needs a human: judging what it found and helping the client further. That's also why every agent in Factor Fifty carries the expertise of someone on our team. AI didn't take Ferris's job. It took the part of her job that never needed her. What's the first task you'd hand to an AI colleague? Let me know in the comments 👇 PS: we all have our own AI agent color at Cloudsisters.
With AI, I think we're still on the rotary phone. 😅 Look at the two black ones. That's bakelite, the first fully synthetic plastic, invented in 1907 by a Belgian 🐓, Leo Baekeland. It then took about 60 years to get to the Nokia 3310, in 2000. From there to the first iPhone: 7 years. 30 years ago, that Nokia didn't even exist. Today we're putting AI in everything, and it can feel like the transformation is already done. I think it has barely started. AI in my work today is that rotary phone. It already does the job, but every week I still see designs that look perfect on screen and fail in the lab. But the door is starting to open. Molecular dynamics adds physics, with no model guessing: we watch whether a design holds up or slowly falls apart before it reaches the bench. So I use what works today, test it at the bench, and keep only what holds up. 😊 Where do you think we are with AI: rotary phone, 3310 or smartphone? Phil. PS: We design antibodies from your antigen sequence. Candidates in 6 weeks, lab-validated in 9.
A delivery manager recently told me that his meetings have become longer since his team started using AI. Everyone comes prepared. The proposals are well researched, the presentations look great, and every recommendation comes with a detailed analysis. But somehow, they spend more time discussing things and less time deciding what to do. I found that interesting. We’ve spent a lot of time encouraging people to make data-driven decisions. Now anyone can produce a convincing analysis, complete with supporting arguments, risks, and alternatives. And I wonder whether we’re creating a new problem. When someone presents an idea, we can ask AI to challenge it. Then we can ask another AI to evaluate that challenge. We can keep going until we have explored every possible scenario. At what point do we have enough information to make a decision? I think this raises some interesting questions for organizations adopting AI. How much analysis is actually useful? Who decides when there’s enough evidence? Are we becoming better at making decisions, or just better at preparing for them? And perhaps most importantly, how much of what we’re trying to figure out could be learned faster by actually doing something? I suspect we’ll need to become much more disciplined about how we use AI in decision-making. Otherwise, we might end up spending all the time AI saves us discussing what to do with it.
++ Why HR should lead AI transformation ++ AI transformation goes far beyond technology. It changes how work is designed, how organizations operate and how people collaborate with intelligent systems. Nadia Shah from Deloitte Belgium explains why HR is uniquely positioned to help shape this transformation and make organizations more adaptable. In this #ZigZagHR BrainPickings short, we explore HR's opportunity to take a leading role in redesigning work. Watch the full conversation with Nadia Shah and Timothy Bruneel on future-proofing HR architecture, part of our four-part podcast series on Deloitte's 2026 Global Human Capital Trends. >> Watch the full episode: lnkd.in/eFgDyw5g >> Watch the full podcast series: lnkd.in/ehBstf3w Deloitte Birgitt Maes
La data et l’IA transforment la santé, c’est entendu. Mais au service du One Health, les défis prennent une autre dimension : il s’agit de faire dialoguer des données qui viennent de mondes différents, humain, animal, environnemental, avec des systèmes qui ne se sont jamais parlé. Le 24 novembre à Namur, cette table ronde réunira le Dr Aloys du Bois d'Aische (Eonix), Eric Christiaens (ZORGI), le Dr Annick Valentin-Smith (Vet IN Tech) et le Dr Fadi Glor (InterSystems) pour en débattre. Vous souhaitez assister à cette table ronde ? Inscrivez-vous ici : lnkd.in/et62pEvt
We're #hiring a new Business Analyst AI in Brussels Metropolitan Area. Apply today or share this post with your network.
𝗗𝗼𝗲𝘀 𝗔𝗜 𝗿𝗲𝗰𝗼𝗺𝗺𝗲𝗻𝗱 𝘆𝗼𝘂𝗿 𝗻𝗲𝘁𝘄𝗼𝗿𝗸 𝘄𝗵𝗲𝗻 𝗲𝗻𝘁𝗲𝗿𝗽𝗿𝗶𝘀𝗲 𝗯𝘂𝘆𝗲𝗿𝘀 𝗹𝗼𝗼𝗸 𝗳𝗼𝗿 𝗰𝗼𝗻𝗻𝗲𝗰𝘁𝗶𝘃𝗶𝘁𝘆 𝘀𝗼𝗹𝘂𝘁𝗶𝗼𝗻𝘀? Your company may offer exactly what buyers need. But when buyers ask AI to help them find a provider, does AI understand what you offer? And does it recommend you over your competitors? We explored these questions in our latest AI Visibility Market Landscape Report, focusing on enterprise connectivity providers in Western Europe. 🔎 We examined how ChatGPT, Claude, and Gemini present providers when buyers ask about site qualification and delivery feasibility, from checking connectivity availability at a new office to assessing high-capacity connections between sites. A few findings stood out: 🔹 Equinix leads in AI Share of Voice (18%) and has the highest recommendation rate (64%). 🔹 Colt Technology Services appears most frequently, featuring in 67% of AI answers. 🔹 Orange Business, Amazon Web Services (AWS), and Digital Realty complete the top five by AI Share of Voice, but their recommendation rates tell different stories. 🔹 Only 19% of citations come from the top five vendors' own websites. The analysis also features Megaport, Vodafone Business, BT Group, Lumen Technologies, and Deutsche Telekom, providing a broader picture of how AI represents the competitive landscape. The results highlight an important distinction: being named by AI doesn't necessarily mean being recommended. 📊 Explore the report to see how these providers compare, which brands AI recommends, and where it finds supporting information. This is a snapshot of one business capability in one regional market. What would a similar analysis reveal about your company's offerings, competitors, and priority markets? Pronovix can develop a dedicated report examining how AI understands and presents what your company offers, whether across a global market or a specific region or market segment. 𝗨𝗻𝗹𝗶𝗸𝗲 𝗔𝗜 𝘃𝗶𝘀𝗶𝗯𝗶𝗹𝗶𝘁𝘆 𝗺𝗼𝗻𝗶𝘁𝗼𝗿𝗶𝗻𝗴 𝘁𝗼𝗼𝗹𝘀 𝘁𝗵𝗮𝘁 𝘁𝗿𝗮𝗰𝗸 𝗺𝗲𝗻𝘁𝗶𝗼𝗻𝘀 𝗮𝗻𝗱 𝗿𝗮𝗻𝗸𝗶𝗻𝗴𝘀, 𝗣𝗿𝗼𝗻𝗼𝘃𝗶𝘅'𝘀 𝗻𝗲𝘄 𝗯𝗲𝗻𝗰𝗵𝗺𝗮𝗿𝗸 𝘀𝗲𝗿𝘃𝗶𝗰𝗲 𝗶𝗻𝘃𝗲𝘀𝘁𝗶𝗴𝗮𝘁𝗲𝘀 𝘄𝗵𝗮𝘁 𝗹𝗶𝗲𝘀 𝗯𝗲𝗵𝗶𝗻𝗱 𝘁𝗵𝗼𝘀𝗲 𝗿𝗲𝘀𝘂𝗹𝘁𝘀. It examines whether AI systems understand what your APIs enable, how your technical offerings connect to your business solutions, and what information gaps may affect how your company is represented and recommended. Reach out to Kristof Van Tomme to start a conversation.
The world of #AI is now!
Where I think AI governance is heading in 2027: This year, most organisations treated AI governance as a policy exercise: write an AI policy, set up a working group, ask people not to paste customer data into chatbots. That was a reasonable start. I don't think it will be enough next year. Three things I expect. 1️⃣ The deadline moved, the work didn't. The AI Omnibus pushed the high-risk obligations for most stand-alone AI systems to December 2027. Some organisations will read that as time to wait. I'd read it as one year to get the basics right: which AI systems you use, who supplies them, and what they're used for. You can't classify what you haven't listed. 2️⃣ AI governance lands with the privacy team, officially or not. The questions are too similar: what data goes in, on what basis, who's accountable, what happens when it goes wrong. Few organisations will build a separate AI governance team. They'll ask the DPO. 3️⃣ AI assessments become part of the same workflow as everything else. Organisations that run AI assessments separately from DPIAs and vendor assessments will end up doing the same work twice. I expect them to merge. I'm curious whether you see it the same way. I'll be at the IAPP Europe Congress in Brussels on 18 and 19 November, at booth #40, and this is one of the conversations I'm hoping to have there.
🔬🧪 TAKE A CLOSER LOOK! Every great scientific discovery starts with curiosity. This National CyberDay, we're zooming in on Artificial Intelligence to uncover the ideas, processes, and possibilities behind this powerful technology! 🤖 There's more than meets the eye when it comes to understanding how AI works! Whether you're an educator looking to better understand AI in the classroom, a student curious to learn more about the technology, or a parent wondering how to start conversations about AI with your kids, we've got resources to help you get started! There's something for everyone. Join us for ICTC-CTIC National CyberDay on November 16! 🧪 Register today at NationalCyberDay.ca and get ready to explore! #NationalCyberDay #NationalCyberDay2026 #CyberDay #CyberSafety #CyberDay2026
Sinds oktober vinden de meeste cursisten de weg naar opleidingen via AI. Afgelopen zomer zijn we catalogi van in Smart Lions aangesloten opleidingsverstrekkers beginnen optimaliseren voor AI-bots. Vorige maand kwam het kantelpunt: AI-botverkeer kreeg de bovenhand op klassieke crawlers en menselijk verkeer. We rapporteren die cijfers bewust niet naar opleidingsverstrekkers, om geen onrealistische verwachtingen te scheppen. AI-bots zijn namelijk abstracte beestjes. Het gaat niet per se om een potentiële cursist die via AI interesse heeft in een cursus. Het enige wat zo'n cijfer wil zeggen, is dat opleidingsinhoud wordt opgehaald door een AI-bot. Puur AI-modeltraining valt daar dus ook onder. Onze cijfers wijzen wel uit dat opleidingen met meer AI-botverkeer ook meer menselijk verkeer verwerken, en effectief ook meer inschrijvingen ontvangen. Geheel onbelangrijk is het dus ook niet.
Talking to people about how AI is changing work brings up so many similarities across roles and organisations. We're all orchestrating more work, with more reviewing and deciding rather than doing every step ourselves. And this is just the beginning, imagine what it'll be like in a year or so. What won't have changed by then, is how good it is to spend time with people, and get to experience novel surroundings. Today's event about AI strategy took place in Kimpton Clocktower, historical Manchester landmark dating back to 1895 when it first opened as a life insurance company. In the photo you can still see the outline of the turning circle used by horse-drawn coaches to bring mail in - and then be turned round to head back out. A bit of trivia: the three metre bronze carriage horse was commissioned by Sophie Dickens, who is Charles Dickens’ great, great granddaughter. The horses don't bring in the mail anymore, but words, exactly like Charles Dickens used to write them, have not lost their power. To the contrary.
Last year, brands bragged about using AI. This year, they're bragging about NOT using it. Levi Strauss & Co. just dropped a stop-motion Christmas ad: lnkd.in/exy3K2WU A six-person team spent 2,000+ hours over six and a half weeks building sets, painting 400 tiny trees, and sculpting four faces for every character. Four hours of work for every single second of animation. Then they published the making-of video next to the ad: lnkd.in/e8mVuDgd That's the real move. The proof of work is now part of the product. Here's what's happening: ➛ The Coca-Cola Company's AI Christmas ad got roasted. McDonald's pulled theirs after the backlash. ➛ Prada went hand-drawn. Apple went stop-motion. ➛ Hermès and Ralph Lauren now credit the humans behind their ads in the captions. Audiences have learned to spot AI polish, and polish alone no longer signals quality. It signals "cheap and fast." So "human-made" has become a premium label, the same way "handmade" or "small batch" did before it. 3 takeaways: 1. Show the process, not just the result. Behind-the-scenes content is no longer a bonus. It's your receipt. 2. Name your people. Crediting the creative team signals craft and builds trust. 3. Use AI where it's invisible, and human craft where it's felt. Efficiency behind the scenes is fine. Your hero moments are where people judge your brand. The question to ask before your next campaign: if someone asked "did a human make this?", could you prove it in 30 seconds? Because soon, your audience will ask. LET'S GET FUZZY | Levi's® Holiday Horror Film youtube.com/watch?v=JLt9A07POhY
Core Super brings home automation, professional video recording and powerful AI together in one future-ready server, with direct KNX connectivity and an integrated high-resolution display. Redundant DC and USB-C power inputs keep Core Super running, even if one power source should fail. A new dimension of intelligence for Basalte Home. 💛 #coresuper #superintelligence #AI #peaceofmind #smarthomeseurity
Your AI agents will shop for you this Black Friday… but at your own risk. The big question of AI liability seems to have an answer. And customers won't like it! Wells Fargo, one of the largest banks in the US, is now taking a clear position. Starting November 24, just 3 days before Black Friday, its new terms state that: - Transactions made by your authorized AI agent can be treated as if you made them yourself. - If your agent makes a mistake or goes beyond your instructions, you may still be responsible. - The bank does not have to verify every transaction made by your agent. And the timing is of course not a coincidence. The rules take effect right before one of the biggest shopping events of the year, as AI agents are starting to enter the shopping experience. That could be the real test for consumer trust in AI shopping. Would you let an AI agent spend your money under these conditions? Source: Wells Fargo Online Access Agreement
Three situations, one question. AI investment, French government debt and the oil price seem to have little in common. For Ilya Vercammen, Chief Strategist at Puilaetco, all three come down to the same thing: how much risk are investors still willing to take, and at what price? In AI, the focus is shifting from growth expectations to the capital that data centres, chips and energy consume. Investors are taking profits in technology and semiconductors, while healthcare, consumer staples and financials draw more interest. In France, investors are demanding a higher return to finance the state. The common thread is credibility. Puilaetco still prefers equities over bonds, but has reduced its exposure to European equities and cut European government bonds back to neutral. “Being constructive does not mean taking risks blindly.” Read the full analysis (in Dutch): lnkd.in/eEekNrww
Is data the new star of the pitch? According to a report published in September 2026 by Bpifrance and Les Meneurs, data, AI and digital platforms are becoming the main drivers of the sports economy, ahead of the traditional model built on TV rights and clubs. At IFREQ, we agree, with one condition: Data only shines when everyone can trust it and understand it. That's why we build a standardised test for football, scores validated by humans, and a pentagon any player can read. The stars are still on the pitch. Data just makes sure more of them get seen. ⚽ 👉 lnkd.in/eKMWh27u Follow IFREQ and join the future of football. 🚀 ⚽ #SportsTech #Football #Data #IFREQ Sport Tech : la donnée est-elle la nouvelle star des terrains ? bigmedia.bpifrance.fr/nos-actualites/sport-tech-la-donnee-est-elle-la-nouvelle-star-des-terrains
We've just DOWNGRADED Claude models Haiku 5.5 is EXCEPTIONAL at retrieving data You have to seriously consider switching: We're testing two things: - How much does it cost for a model to retrieve information from our data - How accurate is the information it retrieved Way we tested: - Each model was tasked with answering 18 questions - Questions were of differing difficulty - We changed each model's 'effort' level (medium, high, max etc) At high effort: - Sonnet = 71.3% accuracy at $5.57 per question - Haiku = 71.7% accuracy at $1.15 per question That's comparable accuracy for ~5x cheaper! We are moving our retrieval agents to Haiku 5.5. Big reminder, keep testing models, even the ones that you think might be worse!
Jusqu’où une entreprise peut elle laisser une IA agir seule ? Toutes les tâches ne présentent pas les mêmes conséquences. Une activité sans effet direct peut être largement automatisée. En revanche, modifier un fichier partagé, envoyer un message, traiter des données personnelles ou influencer une décision exige des règles plus strictes. Avant de connecter ChatGPT à un processus, plusieurs points peuvent être examinés : la nature des données, les droits nécessaires, les conséquences d’une erreur, le contrôle humain obligatoire et la conservation d’une trace des validations. L’enjeu n’est donc pas seulement de maîtriser l’outil. Il faut aussi décider ce qui peut lui être confié, ce qui doit être relu et ce qui doit rester humain. lnkd.in/efjGDmbz #GouvernanceIA #Automatisation #SécuritéDesDonnées #IAEntreprise Dans vos activités, quelles tâches devraient selon vous rester exclusivement sous responsabilité humaine ?
AI is changing the game... and women are leading the way. 💜 Come connect with us at “AI: From Impact to Growth” on 12/11, a special pre-event of The Future of Tech is Female 2026, co-organised with Women in Tech® Belgium, and taking place at BeCentral in Brussels. 🌱 Bringing her expertise to the table, Iris De Coster, representing Proximus Ada, together with Victoria Sarmiento Herrero, Peggy Brione and Nanna-Louise Linde will join the AI & Corporate Transformation panel from 6:40 to 7:10 PM to share perspectives on AI strategy, implementation, business transformation, partnerships, and scaling innovation across industries. 🌟 🤖 Join an evening of fresh ideas and meaningful conversations with inspiring women shaping the future of AI across business, technology, entrepreneurship, and investment. 💡 Secure your spot here : lnkd.in/eES5hv_J Proximus Ada #AI #ThinkPossible #ProximusAda #WomenInTechBelgium #thefutureoftechisfemale
No, I don't 'save 1h a day with AI' 🫠 But it gave me 2 other things that might be more valuable: Everyone is chasing the AI productivity dividend: the four-day week, the "40% time saved", the promise that you'll finish work at 3pm. I use AI intensively every day, and the reality is: I'm not working less so far. You? My day is still packed as it was before, and I’m sure yours is too. But the type of work has changed. 1️⃣ First, AI took a lot of the what I call ‘soul-sucking work’ off my plate. The reformatting, the meeting summaries, the fifth version of the same proposal, the processing of the invoice, and all the other admin that takes time and effort. All little tasks that don't necessarily take hours, but somehow drain your energy. That work hasn't disappeared, but it's no longer mine. 2️⃣ Second, my work got better. I’m better prepared now, I challenge my thinking more, I follow up faster and I forget fewer things. I’m simply just more on top of things, really. And maybe that's the real AI dividend for me. Not fewer working hours, but better hours. What’s your experience?
